Kelly Brown: Why don't we get started? Good afternoon, everyone. Thanks so
much for you joining us this afternoon for Session 2 of this webinar series on
capacity building around demographic data. As folks get in the room and get set up
and get comfortable, just again, want to welcome you to this, the second session of a
series that's been sponsored that is being sponsored by a collaboration of three
foundations, the James Irvine Foundation, the Weingart Foundation and the
California Endowment. If you joined us for the first session, you've got a little bit of
background on the collaboration, which I will share a tiny bit again, just to reinforce.
You will know that we are getting into the meat of what we want to share with you
about strengthening nonprofit capacity to engage, to collect, and use demographic
data. Again, just to remind folks about what the California demographic data
collaboration is, as you can see the three foundations that are listed here and around
the table and whom some of you heard from in more detail at the last session, work
collaboratively in a number of areas, in a number of ways and realized that they had
a lot of overlapping grantees and organizations in common asking.
While they have different priorities, and different focus areas, and different
processes, there really was a lot of overlap and a shared commitment to equity and
justice. They decided to come together to think about ways that they could
streamline their procedures and their information, the data that they request from
nonprofits, particularly since there was so much overlap, to not only ease the burden
that nonprofits faced with reporting to so many different kinds of entities but also to
strengthen the capacity to make better use of the data because what we learned was
when each individual foundation or entity collects data from nonprofits that data is
often quite siloed and proprietary. It's not really used to its full capacity.
That was the impetus for the collaborative. What we leaned into was really trying to
center nonprofits and their experiences and capacities, and collecting this data, and
really look for strategies and tools and resources that might strengthen that capacity.
That was the goal of the collaborative, and that's how this webinar series started to
unfold. As I mentioned we scheduled five in this series. We started on April 25th with
a emphasis on why this data is so important. Today we want to talk to you about the
what. Like, what are you collecting, and why is that related to what is important, and
how you're going to use it?
We're going to talk about taxonomies and terms and give you a little bit of history,
and some key pointers. While I will be talking a bit more over the course of the next
few minutes, I will be joined by my colleague, Audris Campbell, who is the founder of
Research Gurus and is a survey researcher and methodologist for many years. She
worked at Gallup and IST Research, and formally with Google. She brings a wealth
of experience with just good data collection and good tools to bring to this table.
She's really been an asset to the team over the past couple of years.
We'll turn it over to Audris toward the end. Then we can sum up and get ready for the
Completed: 08/28/2024

1

next session. I also wanted to share and remind folks that the sessions are going to
be about 50 minutes. We have a little bit of leeway at the end in case you went over,
but we're really trying to keep it to 50 minutes. We heard in your feedback at the end,
from the survey that we shared with you in the first session, asking for how can we
improve the session, what are the questions that are burning for you about these
kinds of issues? We made some refinements to the process to deliberately and
explicitly pause and allow for people to engage the Q&A and to ask questions.
We really encourage you, as I go through the first part of the session, to think about
questions that you may have, that may surface. This is not in any way to preclude
you from jumping in if you have a question, but we're really going to create some
space to make this a little bit more active. We'll save time for questions again at the
end. Just to remind folks that these webinars are being recorded and they will be
supplemented with curated materials and information that doesn't replicate but
reinforces and supports what's going on in the deck, and all of that will be shared at
the end of the series.
If you miss something or you want to go back to something, you'll be able to have
access to that. We're going to start to cover a lot of ground today in a short period of
time. I just want to outline that the issues that we'll be talking about in terms of how
do you classify people, what are the terms, what are the frameworks, that's actually
what a taxonomy is, is representative of a history of dynamism, and inclusion. We'll
give you a little bit of that history, but really want you to know that these terms and
concepts and terminology and phrasing are actively and robustly being engaged by
researchers across the country and really, across the world.
I'm actually in Philadelphia right now at a conference of survey methodologists where
these topics of how do you classify people and why, how do you get the best
information, is really a hot topic. It's going to be a lot to condense into a short period
of time. It won't answer all of your questions, but it's just designed to give you a
sense of the reasons why this is so complex and yet still so, so important. It's really
going beyond the data, beyond deeper than these numbers and the checklist, to
really understand the complexity that is underneath identity and why that matters.
Then, as I mentioned, we'll end with some concrete tips on getting good data and
seeing people. In other words, getting good sound data, from as many people as
possible while also trying to be as inclusive as many of you need to be, particularly in
your areas and context. I want to start with just, again, a reminder that a taxonomy,
often confused with a survey instrument, is really just a way of classifying things. It's
a classification system. It's, how do you identify characteristics, in this case, of
individuals, and put them into groups of it? You can make meaning of what those
characteristics have in terms of your setting and your context.
It's also important, though, to remember when you're asking the questions and
framing the tools, that it's important to be clear on what you actually need to know to
inform action, as opposed to what would be nice to know. Many times people go
from extremes from asking nothing and not knowing much about their stakeholders
Completed: 08/28/2024

2

and constituents to asking many, many questions of all types that are also
burdensome to individuals and may or may not be insightful or relevant, but which
may just be routine. We're going to try and give you some guidance to how to
navigate that and how to balance it.
We do want to emphasize, though, is the more complex questions you ask, it
produces more complex data that then needs to be analyzed. We want to avoid
situations where people are asking lots of data and then not really being able or
capable of using it, so, balancing that actionable data with inclusion. I'm going to stop
and pause here for a moment, as we did the last time, to see what are you collecting.
From the last conversation, we saw that the majority of you were engaged in
demographic data, in some way, on your stakeholders, your leadership, your staff.
What we want to do this time is ask what kinds of data, what kind of demographic
data do you collect? We've listed here, some common ones that people tend to
collect, but we also know that folks collect other things. If you can go and think about
the data that you collect, particularly on your stakeholders, constituents, end users,
in other words, people who are outside of your organization, what are the
parameters that mean something to you? I would ask, I see a number of folks in the
Something Else box, if you can put in the chat what that something else is.
That would be super helpful because then we can see what else is of interest to
folks. Just drop it in the chat. Again, it's not designed to be scientific, but it is
designed to give some indication of the range of things, characteristics, et cetera, of
interest to folks. As we're getting more people replying, some of the responses and
the weighting of the responses is actually common in what we see in other settings
and across the board is that a lot of folks are collecting and focused on race and
ethnicity, and in many cases, gender. Folks are putting in the chat, immigrant age,
languages, all of those are very common.
What we also see is much considerably less collection of sexual orientation and
ability. Of course, that's for a combination of reasons. Over the course of the series,
we'll engage the applications of some of those parameters, but this is very common.
When you think about your constituents, these are parameters that you need to
know. We got about 76% participated. It looks like it's slowing down, so that's super
helpful. Let's see, just share the results so folks can see. Again, as we said, it's
mostly race and ethnicity. Not far behind, gender and identity, sexual orientation,
ability, even less so.
Something else in the chat is showing is age, immigrant status, language, what have
you. The next slide, I want to give you yet another poll, if you will, real quick, not to
belabor this, but we do want to ask, what do you collect on your staff and
leadership? That is slightly different from what people are collecting on their
constituents. I'm having a little trouble getting the poll to launch. It keeps showing me
the results. Oh, here we go. Here's the poll for your staff and leadership. Again, a
different set of stakeholders, similar kinds of parameters.
Completed: 08/28/2024

3

What, if any, data do you collect on those stakeholders in your organization? In other
words, what is it about yourself and your organization that you collect? As we go
through, not only the session today, but also in subsequent sessions, we will talk a
little bit about what is the relationship between the characteristics of your staff and
leadership, and your parameters. Anything else in the chat in terms of something
else that you share on your leadership, years at the institution, that's super helpful. Is
it, are they representative potentially of the stakeholders that you're serving?
It also looks like there's a lesser number of folks, potentially, who are collecting
information on their own organization. We'll talk a little bit about the implications of
that. I'm going to end the poll and share the results real quick so people can cut
stats. You see, again, a little bit of alignment, mostly race and ethnicity, gender
identity, less so sexual orientation, and ability, and in some cases, something else. I
wanted to put this out there so that people can see where is the energy in terms of
these demographic parameters. Often it is around the hot topics of race and gender,
less so sexual orientation, but actually, what we know in California, it's actually a little
bit more common for people to engage around sexual identity and sexual orientation.
We'll talk about how a lot of these parameters and terms depend on generational
factors as well. Taxonomies, in general, as you know, I call it the tyranny of the
taxonomy because it really is a classification system that has been going on since
the beginning of the country when people had very simple parameters around which
they collected data on. As you can see from this graphic over the years, it has gotten
more and more granular and more and more complex. As the nature of the country
and the demographics of the country have changed, so has the classifications and
the ways in which the country is collecting information on who is existing.
As you may be aware, there are ongoing and regular conversations in the larger
public about more and more insight and awareness of the complex diversity of the
country. What we're not seeing are tendencies or trends toward being all the same or
one big melting pot. We're increasingly in a space where people want to be seen for
who they are. Having said that, the level of granularity that often is people want to
engage when thinking about the mechanism and the terms that you might collect the
data on, it's really important to strike that balance.
This is something that we will be going back to and reinforcing again and again, is
that the terms that you use can often sometimes be very specific to populations,
sometimes to demographics, to constituencies. For example, in this example from
Grantmakers in the Arts, about their terms that they were proposing using for gender
identity, includes terms that were probably very irrelevant to their own stakeholders
and constituents. That is critical to see who folks are and making sure everyone is
seen. This is not the kind of survey, or they're not the kind of terminology that might
play well in a larger, what we would call general population, where there are lots of
diverse folks in your population who might be experiencing these questions and
queries because in many cases, for a general population, this would be like getting a
survey in another language.
Completed: 08/28/2024

4

Because surveys, people spend so little time on these kinds of queries, you really
want a tool that both captures a range of folks, but also allows for people to
comprehend what you're asking. This is also the case and often comes up when
people think about doing work in international settings or in contexts outside of the
US. Often people struggle with trying to map concepts and ideas of identity that are
happening in other parts of the world, with those same concepts in your local, in your
domestic space. In this case, in the United States.
That really isn't feasible. There really isn't ever going to be the capacity to map the
whole world onto the American concept of race and ethnicity, or any other
parameter, even gender. That complexity doesn't mean it's not important or that we
shouldn't try, it just means that the work has to be done to really understand those
concepts in those contexts. For example, there was a recent article in The
Washington Post around how India is struggling with how to design its own census
questions with respect to caste because caste, in India, has implications similar to
ethnic identities and disparities in other parts of the world.
That's why they need the data to address that and redress that. Similarly, in a place
like South Africa, where one might assume that there are similar kinds of racial
differences and constraints, and histories, how South Africans really define and
relate to the categories of how people are structured and to the degree to which
those categories and foreign policies are really different than what is happening in
the United States and would not map well, the terms would not map well here and
vice versa. I wanted to share this one framework with folks, however, around why,
particularly with respect to race and ethnicity, it's so challenging to think about how to
get the terms and the concepts right.
When you think about race and ethnicity, you're really asking about three aspects of
identity that are important and interrelated, but you're trying to get it out of one
question. In many cases, people are thinking about the phenotype, how people are
perceived, or how they are engaged by the world, what researchers call street race.
You're also, in some cases, trying to engage nationality, which we know is a crucial
factor in terms of identity and makeup, but it relates more to country of origin or place
of origin, and language. Then sometimes you're asking, you want to get at ethnicity,
what are the cultures, practices, and norms, habits, values that people have that are
associated, in many cases, with their country of origin or with their appearance in the
world, or their manifestation in the world.
They are really distinctive but interrelated concepts that people are trying to get at in
one fell swoop. I wanted to share this because I think it helps people understand why
getting at some of these concepts and the terms that you can use to get at them can
be so complicated. I want to also share with you the strain, the tension between,
again, using these terms to see everybody. Though these word clouds are indicative
of what people's individual responses to a survey about, on the left, race, ethnicity,
on the right, sexual orientation of a survey that was recently launched a couple of
years ago by a philanthropic affinity group.
Completed: 08/28/2024

5

These were the write-in responses to a list of options that were presented to folks. In
other words, this is what people wrote in, in terms of how they wanted to be identified
by race and ethnicity. Obviously, the larger the word, the more people gave that
answer. Similarly, on the other side with sexual orientation, the larger word, the more
people gave that answer. As you can see, there were lots of other words and ways
that people really wanted to be seen in these contexts. When you're designing a tool
that is trying to, again, be inclusive and see everybody, you would generally have to
have a tool that had all of these options, which is not really feasible.
However, it is important in these contexts and in these tools to allow people the
option to self-identify in some other way. The term that is emerging is, "I prefer to
identify as something else or in some other way." It used to be other, and people felt
that was dismissive. We are in a moment where that other, that preference to identify
in some other way option is really an opportunity to learn about how your
communities and constituencies are changing, and the terminologies that they are
evolving and growing into. Again, some terms are repurposed, some terms are
revisited and reclaimed.
That option for people to choose can really give you that insight. We just think that
folks want to pause and think about what are you already collecting in terms of the
data. We saw folks are largely engaging race and ethnicity. A lot of folks are
engaging other things. The difference between intrinsic characteristics, those that
folks carry with you that are often visible, that are generally immutable, versus those
that are acquired as people move through the world and are moving in different
social contexts. Important to think about how do you get data that gives you the
information that you want to know, and that helps you understand the intersection
between these two elements.
Just to give you some examples, so when you're thinking about, for example, a
tenant's rights organization, a direct service organization, your goals, if you focus on
what are you actually trying to change, what are you trying to impact in the world, in
this case, housing stability, reduced evictions, what is the kind of demographics that
may influence those outcomes or the field in general? It could be a long list. It could
be lots of things that you anticipate may or may not influence. Then when you focus
on who are you intending to serve or engage, that can help you focus more on how
to hone in what is the taxonomy that you're actually really going to use and design.
Again, what we learn and what we recommend, and probably what many of you are
already finding, is that you may begin with some hypotheses or some suppositions
around what's important to know. As you go out and get information from individuals
and folks, you will find that some information is more important than others, and it's
okay to start to refine and reduce and streamline what is the data that's really key.
Having said that, it's also important, particularly for organizations that may be
national scope, that may be more advocacy-oriented, that may be trying to reach the
populations at large or the general population.
In many cases, folks are assuming that the demographic characteristics that they
Completed: 08/28/2024

6

know can be pretty straightforward or generic in terms of the demographics of the
country as a whole. It can be really important to dig a little bit deeper. Are there
variations geographically? Are there variations by age? Are there variations by party
that may inform this? When you think about COVID, for example, or messages about
COVID, the country, at the national level, had a primary messenger with a primary
message around the vaccine and reducing risks.
We learned over time, as the pandemic unfolded, that the messengers and the
messages really varied a lot by community. It was really important and critical to
have that data and that information, even when you're designing strategies that are
intended to reach the whole population. I'm going to stop here and see what the
questions are. I see some things in the chat, but would love to hear if there are other
questions that people want to drop in for surface.
Meghan McVety: While we're waiting for that, Kelly, I think what you're speaking to,
there are a couple of comments about recognizing that demographics are key to
understanding your strategy at the beginning of designing your strategy. it's not just
about collecting data at the end of it. That seems to be something that's coming true
and people are doing as well.
Kelly: Exactly. Again, I'm glad you somewhat pointed that out and why we began
with the why, the, what are you going to do with this data? What can you possibly
change? How are you going to think about your services and programs? It's often, I
won't say too late, but people often do count up what they've already done. That's
sometimes after you could have designed a different sort of intervention. What we
would argue is it's never too late to really get that information and be open to
challenging some of your assumptions. I think one important thing to resist, which
often happens with this data is people are anxious that it will tell them that they're not
doing something right for certain groups when really, people should be leaning into
thinking about how can they get better, and how can they improve.
Meghan: We've got a couple of questions here. One from Hanh. Hanh, do you want
to speak here since you're here?
Hanh Cao Yu: Sure. I really like the fact that you tied data to the work that you're
doing, your theory of change. What might be some reasons to collect certain
demographic datas even if they don't relate to programming or strategy?
Kelly: Right. Again, it's interesting to think about what people may have assumptions
for about what is the demographic data that's relevant or not relevant. For example, I
think in the last session we talked about an organization that was a women's
reproductive health organization. Come to find out-- One would think that maybe
gender wasn't something that you needed to probe on because it's a women's
reproductive health organization. When they did ask those questions, they found that
there was a lot of gender variation and that men were actually participating in their
programs. It may not seem obvious on its face that you need this kind of data for
your programming, but what we are learning, and what I think generally in the field of
Completed: 08/28/2024

7

social science, in politics, in health, in other arenas, the parameters around race,
ethnicity, gender identity, to some degree sexual orientation, certainly ability, but not
exclusively those parameters, in some cases social, economic, have been known to
be factors that really vary a lot, vary consistently over lots of different interventions
and lots of engagement.
Minimally, those are the kinds of things that people should be probing, but once you
start to get that kind of information into your organization, it will start to uncover
potentially other kinds of demographics that might be more salient. Are older men
more likely to engage our program than younger men, vice versa?
Meghan: We have a few other comments, questions. Francis would love to hear
more about the balance of inclusivity being seen. I know you're going to address
that, you and Audris, so I'm just a bit on taxonomy tips. Francis, just know that that's
coming. A comment, "How people define identity is often much more variable than
the standard taxonomies." I think too, you'll get into that as well. Appreciate that
comment also about the tenant right advocate example, which I think really reflects
the range of organizations and why data is needed. That's great too, and know that
we'll be building on this as we talk about collection strategies in the next session and
putting the data to use as well.
Kelly: Yes. Those are really excellent thoughts and comments from folks. I think the
point around how people identify as a variable is really actually super critical. That
was why I tried to share that framework around all the different aspects of identity.
Those aren't the only ones, but those are the ways, and they're variable. In other
words, they're finding that people actually change over time in terms of how they
think about each other. That's why the ability to allow individuals to self-identify if
they want to or choose to, can surface a lot of that variability.
Over time, in many cases, it does begin to aggregate up into the larger category.
People are talking about having only limited options to put people in. As we're
saying, just in the past census and going forward, several new options such as
Middle Eastern and North African are emerging. There's conversations about making
the Asian category more granular, and you could see that over time, so the graphic
at the census level, that is the direction that people are moving in. It does take time
because you're having to balance that evolving sense of who people are and how
they're defining themselves with still understanding, in general, what are the
parameters of the population and having data that you can then act on.
Meghan: Yes, and one more thing before you move on is just to acknowledge the
frustration that there can be with the variety of ways that data is requested. Just
know that the three foundations have come together to exactly address that and try
to streamline and standardize as part of a movement that is growing. I want to
acknowledge that we're in the middle of that process, or we hope to be.
Kelly: Exactly. I can hear the frustration coming about the balancing. Some folks
who want to see every single parameter, and we'll talk about how do you actually
Completed: 08/28/2024

8

manage that data with grouping things into the larger categories. It is an ongoing
tension, but again, to go back to the impetus for this collaborative was how do we
start to move toward some kind of alignment while still building tools, and
mechanisms, and processes? Part of that process is engagement directly with folks
that allow for mutability and dynamism over time. I also see a question about what
foundations are doing with this data that they ask.
I'm glad you mentioned that because I was going to mention this at the end, but I can
surface it now. The final session that is billed as the town hall where we can have
more of this robust discussion between all of you across some of these tensions and
these concerns will be an opportunity for you to talk directly with the representatives
from the three collaboratives. To the degree that they are and are not willing to
speak on behalf of their peers, they can at least tell you what they're going to do. On
that note, I want to get Audris's points in terms of some good steps around trying to
navigate some of these. Then we will leave open some time for some additional
questions and go from there. Audris is going to give you some tips on how to design
your taxonomy and how to, as we said, navigate some of these complexities that
you're raising.
Audris Campbell: Awesome. Thanks, Kelly. Let's get started with a few tips. The
first one is reflecting a little bit about what we're hearing in the comments, but
consider using an established taxonomy. Intentionally aligning with an established
taxonomy has a lot of benefits. One of those is that demographic experts have
already consulted on the categories because often, the first step is what do we say
and what do we use? How about we just use something that the experts have
already agreed on? That's one great thing. Another thing is you can compare your
data with the field.
You can see what others are doing and how what your data says compared to
others, which is great, and see how you measure up. Then this third point here,
which is also being talked about in the comments, is just reducing burden. The more
aligned we are, the better it is. We know that a lot of nonprofit organizations are
reporting to a lot of groups, and also, a lot of foundations are collecting a lot of data,
and it's often not aligned. By using established taxonomies and standardizing, it
helps everyone. Next slide. This gets to something Kelly was talking about earlier
when she was just mentioning that taxonomies and terminologies change over time.
My second tip is to stay current on the terminology, but know that it might change.
There's this bit of knowing what's going on, but also being somewhat flexible. Over
time, as we've seen throughout history, the way we talk about identities and groups,
and experiences have consistently changed, and it's not going to stop. It's going to
keep changing. Something that we can do about that is just know what's going on,
but determine when it's a time to make that change and when it's not, which I think
can be difficult. A few things that we would like to talk about is one, updating your
demographic data on a regular basis.
Instead of every time the wind blows, changing everything, have a regular cadence
Completed: 08/28/2024

9

to your demographic update. Maybe that's every two years, once a year, whatever
your organization decides. Just have a regular time basis that you update it so you
don't feel this scramble to change all the time. The other thing to do is think about, at
the bottom I have here, keep in mind, think about who you're surveying. Normally, or
who you're asking these questions to, when you're doing something to the general
population, which means basically, general constituents, not a subpopulation, be
cautious when using terms that are not broadly understood.
Often, terms will emerge in subgroups, and then they'll, at some point, filter out, but it
may take a while before everyone understands the new terminology. When you're
saying, "Is it time to make a change?" think about who you're surveying or who
you're collecting data on. If it's general population, think more about, will everyone
understand for this particular effort. If people don't understand, they might not
answer it all together because they feel lost. However, if you're doing something in a
subgroup, then, as Kelly showed earlier with the Grantmaker's example, that might
be a time to be a little bit more specific because that subgroup might understand
those terms.
Then the last thing I want to cover on this is the struggle with changing terminology,
and why you don't want to always do it is because you have to map responses. If
you're looking to trend your data over time, to look at how things improve or decline,
every time you make changes, it impacts your trends. The ability to say that things
change over time really is impacted. Sometimes it doesn't measure up perfectly, so
you add this new category, now you don't know what to do with this one. That's
another thing to think about. When you do decide to make the change, as closely as
possible, try to map those responses.
How can they relate in a way that you can make sense of your progress or declines
that may have happened. Next slide. The next thing here is keep it simple, but
capture detail when you need it. This is also emerging in the chat, too. I know it
sounds contradictory. We're like, "Keep it simple, but capture detail when you need
it," but you can do this. What you want to do is you want to use broad, high-level
categories. Someone mentioned that their organization has all these categories, but
when it's time to report, they only need to report five. What you want to do, going
back to that mapping, is figuring out how some of those categories roll up to what
you have to report out.
Those would be your broad, high-level categories. If you're collecting this online, you
can have the option to, if someone selects this, then it opens up to smaller questions
where you can get more niche and detailed. That way you have the broad categories
and the data that you need, but you can still have these roll-up categories that you
need when you're interested in specific groups. I know this might be relevant,
particularly in a state like California, where we know there's so many different
groups, so many identities. When you look at larger United States of America, they
may not get that detailed.
You may be reporting to someone who wants something very broad, but then you
Completed: 08/28/2024

10

need something detailed for yourself. This allows you to capture the internal data
that you need, but also to easily report out categories. Next slide. The next thing we
want to talk about is to consider basically asking race and ethnicity together, and
also consider using Mark all that apply. I'll talk a little bit about race and ethnicity first.
The US Census has been doing a lot of research, and they found that when we
combine race and ethnicity together, it improves data quality, and particularly for
those who identify with the Hispanic or Latino identity.
Generally speaking, when we see race, that is one question. Ethnicity just talks
about, normally in America, Hispanic identity. What ends up happening is when we
combine it together instead of asking it separately, we find that fewer people skip
questions, and also fewer people select Other because race and ethnicity is
sometimes not as separate as we might think they are. For a lot of individuals,
they're actually a little bit more murky, and so we get better data when we put race
and ethnicity together instead of just separating it apart.
Then in terms of Mark all that apply, this also relates to something in the comments,
which is, how do we make sure people are seen, but we're still getting the data that
we need? Mark all that apply is really something that helps those who may identify
as multiracial or have a variety of backgrounds. Even as we go back generations, we
might have different ethnic identities. Sometimes this allows individuals to be able to
express the range of identities that they see themselves. However, this does add
complexity. Some things to think about is when you do combine race and ethnicity
together, it can be a challenge when you compare this to two separate questions.
How do you compare two separate questions to one? You may need someone who
understands data analysis a little bit, how to make two questions one and
comparable. The other thing to think about is when you do Mark all that apply, you'll
have to do a little bit more work with reporting. If the person you're reporting to says
multiracial, but I have Mark all that apply, then I have to figure out which ones are
multiracial. Then how do I take those away so I can get the actual true percentage?
It does increase complexity. That is something that we want to keep in mind.
Even though some of these are best practices, you may want to consider what your
capacity is as an organization and what you're able to take on. Next slide. Then this
one here is particularly just talking about separating out sexual orientation, gender
identity, and transgender, particularly when you're hoping to measure different
things. Sometimes on surveys, we might just see a question that says, "Do you
identify with the LGBTQ+ community?" Which, at first glance, you might say, "Oh,
this is a fine question." The issue with that question can be that oftentimes that
acronym captures a lot of different things.
We're talking about sexual orientation in that acronym. We're also talking about
gender identity, and we're also talking about transgender status. It's so much going
on within the acronym that if you want to get more quality data on the specific issues,
it's best to separate it out and ask them separately. There is a trend moving towards
asking about one's sexual orientation, then asking about one's gender identity, and
Completed: 08/28/2024

11

then transgender status. Oftentimes we'll see sexual orientation and transgender
looped together, but there has been research that found that individuals who are
transgender, I think it's a little bit over 60% will often identify as heterosexual or
straight.
Depending on what you're trying to measure, if you want the best quality data, it's
often best to separate them out. Next slide. The next tip we just want to talk about is
making sure that we include a Decline to state or Choose not to disclose option for
every question. Even though we're sharing all these things, we understand that there
are going to be some individuals who, for their own reasons, decide that they don't
want to report this information. One thing in the research community or collecting
data community, that demographic data should always be voluntary at the individual
level.
What we mean by the individual level is an organization will ask it, but individuals
should still have that option to choose not to disclose. If you're required to report, you
have to report, but you can give individuals a space to say, "I choose not to
disclose." We always want to give them the option to opt-out if this is not something
they want to say. That's also good data. Knowing that someone doesn't want to
report is also good data. You want to know how many people in your organization
don't want to report. As it just says here, some individuals, even if the survey is
anonymous because there's this idea that if we make it anonymous, then everyone is
going to participate, but even if it's anonymous, some individuals still don't want to
participate, and they still don't want to disclose.
Giving them the option not to disclose is very helpful. As I said, if you're required to
report, when you report out your results, you can just include a box in there that
about 25% of individuals chose not to disclose or not to report. As I said, that's great
data for everyone to know who's not interested in responding, but it's just great
research ethics for individuals in allowing them to opt out too. Next slide. This is just
an example of balancing this idea of less detail and more detail. I'll pass it back to
Kelly after this, so if she wants to add on some more.
As you can see, one example is a little bit less detail, still kind of full, but less detail.
You'll see the more detailed category when it comes to Asian American/Asian,
there's these roll-up categories. This is where you're able to roll up. You have that
general, so you can report what you need, but for your own organization, you can
have the detail that you need. As I mentioned, if you are able to collect this
information online, it can look a little cleaner because then you could just have an
option that if someone selects Asian American/Asian, then it opens up these other
categories. Everyone doesn't have to see that and it doesn't have to be as messy. I'll
turn it back over to Kelly at this time.
Kelly: Thanks, Audris. That was super helpful and full and hope folks will get the
deck. I think I wanted to make a couple of points before we started to wrap up and
again, and see if there are any final questions before we prep for next time. That is,
there is a very real tension, and we're trying to give people the support to help
Completed: 08/28/2024

12

navigate that. Granularity versus aggregate, these are tensions that people and
demographers are really trying to navigate as people themselves evolve and as
communities grow and change. That's the journey that we're all on, which is why it's
so important to engage and to understand.
Having said that, for example, the race and ethnicity category, as she had
mentioned, in race/ethnicity, it's helpful for the Hispanic population or Hispanic. Race
and ethnicity is really more salient for the Middle Eastern, North African communities
because their notion of race and ethnicity are different. There really isn't a right
answer with that. It's not either and or, but depending on the context, it's both. To the
point also that was in the chat around, what are people requesting, foundations, and
others requesting with respect to a level of granularity, I think this is where there's a
real opportunity to have good conversations about what do people actually need?
Just when we started it early on, what are people going to actually do with this level
of granularity? On the one hand, when I gave the example about the Grantmakers,
certainly there was some reason to understand people in that fullness, but
sometimes it's knowing all of the different kind of multiracials may or may not be
helpful. It may be helpful, but it may not. It may be interesting, but it may not be
helpful. In other words, it may not be data that you're actually going to use. Our
foundations and others requesting information that is interesting and perhaps
insightful, but may not actually lead to strategies or actions, at least at any given
point in time.
I just wanted to point those out before we move into the sum-up and open it up to the
chat and see what are your take- aways from this conversation that went very
quickly. Let's see what folks have to say. I saw something in the chat there. Meghan,
is there?
Meghan: Yes. We have a great recommendation or affirmation that foundations
should include a statement in their request explaining why they asked for data and
how they use it. There is a lot of mystery in foundations, and we know that, and I
know these three foundations are actively trying to break that down.
Kelly: Also to break down, as Audris was referring to, and others referring to, is
there is this tension around reporting and knowing, reporting and understanding. On
some level, reporting is important. We just had a conversation earlier today with
someone who does a lot of work with HUD. They're very, very, very data-driven and
very, very granular, but they use that data to inform their strategies. They need that
granularity on individuals. In many cases, folks really need to understand what is the
impetus around reporting. Reporting to whom and for what and to what end, really
needs to be a conversation between those who are requesting and those are
providing because it could be that organizations like yourselves who are on this
table, need a certain level of granularity, maybe more or less than the entities that
you report this data to.
I think there's an opportunity, and hopefully, we can do this as we move forward.
Completed: 08/28/2024

13

Let's start that conversation so that the alignment that we're trying to work for is
alignment for everybody. Not just alignment among foundations, but alignment within
the sector.
Meghan: Yes. We have another comment from Francis, and you're great at segues,
recommendations for easing the burden on answering for a group. That will be
covered in collection strategies in our next session. Thank you for that. I'm also going
to put our survey in the chat. It's just a one-minute survey to give us some feedback,
which we'd appreciate.
Kelly: Yes, we really appreciated the feedback that folks gave from the last survey,
and based on the nature of the content, try to, again, make the sessions a little bit
more active and a little bit more relevant. They are feeding on each other, and in the
next session, two weeks from now, we are going to talk a little bit more about, once
you have a sense of the terms and a sense of the granularity, how do you actually
get the data? How do you talk to people about it? How do you convince them to give
it to you? The difference between survey data, aggregate survey data, and
administrative data, privacy issues, storing it, things like that, so that people
understand that while it seems onerous and burden, and can be upfront in the
beginning, it really is something that's manageable.
It's manageable particularly if people understand how they're going to use it and why
they need it. The next session is on the 23rd of May. Oops. after that, we'll have a
session on June 6th where we're more and deeply engaged in using the data,
probably with some case studies and some examples. We hope to give you a little bit
of a high-level oversight on, what are some of the legal issues. What are the things
you can't use the data for or shouldn't use the data for? Then, as I mentioned, we'll
wrap up in June with the open questions for each other and for the foundation
parameters.
On that note, we're a little 1 minute over our 50 minutes. We really appreciate folks
coming and attending and hanging in there. We really ask you to fill out the survey,
give us your feedback so we can continue to refine. We will be sharing the
information at the end of the series, but we're really trying to curate it as we go along
so it answers as much of your questions as possible. On that note, I'm going to stop
sharing, and I'm going to say thanks to everyone and see you in two weeks.
[00:54:22] [END OF AUDIO]